Residential electricity consumption is continuously increasing and accounts now for about one third of the total electrical energy produced in Europe and the U.S. How much residential electricity is used depends primarily on the operated household appliances and the behavior of the residents. One major difficulty for individuals who are interested in saving energy in their household is the lack of information about their electricity consumption. Feedback on energy usage is typically only provided by a monthly (if not yearly) utility bill and thus remains rather vague and opaque to most residents. As a result, most individuals could reduce their electricity consumption, but few know how much they consume and even fewer know how much energy they consume for a particular purpose (e.g., lighting). And even those who do have a fair understanding of their consumption patterns rarely receive guidance about the changes that will have the biggest impact on their electricity bill. Through recent technological advances in terms of cost, size, and computing power, Information and Communication Technology can help in many ways to address the challenge of making residential electricity consumption visible to individuals. Embedding computing and communication devices in everyday objects, as advocated by Ubiquitous Computing, can help to communicate the consumption, but also the most energy-efficient usage of a particular smart appliance. Smart meters that enable capturing fine-grained electricity consumption information at high frequency are currently replacing traditional electricity meters. Smartphones have become ubiquitous, powerful computing platforms that allow visualizing energy consumption on the spot without the need for external wall displays. By digitally enhancing physical devices that populate homes, Ubiquitous Computing is offering new possibilities to address the problem of residential energy conservation. Applying Ubiquitous Computing technologies for residential energy conservation raises research questions about the most suitable overall system design of energy feedback solutions and the most appropriate modality of communicating the consumption and guidance information to the consumer. This thesis addresses these research questions by examining how Ubiquitous Computing can help provide effective feedback that goes beyond mere consumption values and is at the same time integrated into daily life. Following a user-centric approach that combines the use of smartphones and smart meters, we tackle some of the open challenges in residential energy conservation. The contributions of this thesis are threefold. First, we design, develop, and evaluate an electricity sensing and feedback infrastructure that seamlessly integrates into the residential environment. It addresses the technical requirements that have been identified in previous research to enable users to better understand their energy consumption (i.e., integration into daily life, real-time information provisioning, low usage barrier, and fine-grained consumption information). At the same time, the infrastructure serves as an easily extendible framework that can be used by other researchers (e.g., to develop and test visualization concepts, to realize further automated energy savings, or to design behavioral science experiments). To demonstrate the feasibility of our approach, we implemented a prototype of the infrastructure and deployed and evaluated it in a laboratory setting as well as in four households in Switzerland. The architecture supports the interaction capabilities of mobile phones together with the integration of smart electricity meters and is used as the base for most other work done in the context of this thesis. Second, we evaluate the potential of mobile phones to serve as portable electricity feedback monitors in two different experimental settings: a user study as well as a real-world deployment. In the user study, we analyze the perceived value of various feedback functionalities and identify which type of feedback is meaningful to users. Moreover, we evaluate the general usability, accuracy, and intention of use of such an electricity feedback application. The real-world deployment aims at characterizing different user types and providing qualitative results gathered through the use of the application. It shows that to foster long-term application of the system motivational concepts are required that engage users once their initial curiosity is satisfied. Overall, the results confirm the suitability of mobile phones as an energy feedback interface and provide insights for the design of future energy conservation applications. They outline that a clear and easy to explain use case scenario is key and that knowledge-increasing functionalities as well as those functionalities from which monetary savings can be directly implied are perceived as most important. To address technophobe users, action-guiding feedback that goes beyond displaying aggregated information in mere numbers is required. Third, we develop, implement, and evaluate an algorithm that disaggregates the overall energy consumption to the consumption of individual devices. It enables users to link consumption with behavior and provides the base for automated energy recommendation systems. Compared to other load disaggregation approaches, our algorithm does not require additional hardware nor complex, time-intense calibration conducted by domain experts. Moreover, our approach is able to easily take new appliances into account where other systems require recalibration. With a simple yet powerful feature provided by the user interface on the mobile phone, users can incrementally integrate additional appliances into the disaggregation process. This is particularly important in a fast changing home environment. We evaluated the performance of our system in a laboratory test study with eight simultaneously running devices, achieving recognition rates of almost 90%.